INDEX

Explanations

alias and aka

The neuron strongly activates on stylized alias markers (forms of “a.k.a.”, “AKA”, “alias,” or variants like “a / k / a”).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 jeżeli

-1.33

 itemName

-1.31

достатки

-1.29

when

-1.27

can

-1.23

 funcionar

-1.18

って何

-1.18

 include

-1.16

</h4>

-1.15

POSITIVE LOGITS

ingang

1.41

kipun

1.33

[];

1.33



1.32

 Новости

1.30

 miatt

1.29

 diskusi

1.28

NCF

1.27

REFERRED

1.24

 DEAD

1.23

Activations Density 0.044%